Multi-view point cloud scenes mosaic is an effective method to solve the incomplete object data problem while self-occlusion and occlusion happened in ladar three dimensional imaging process. Mosaic method directly affects the object detection and recognition. By using projective distribution entropy, the independent coordination of point cloud scene is estimated. Based on this, space transform parameters between multi-view point cloud scenes are calculated. Then, the annealing interative closet point (ICP) algorithm is used for precise mosaic. Experimental results demonstrate the effectiveness and feasibility of the proposed algorithm.